Pricing

The Pricing page lists the models Code Router exposes through its OpenAI-compatible gateway and the billing rate that applies to each one. Because Code Router is a relay that fronts many upstream providers behind a single API key and base URL, the rate shown here reflects what your deployment charges for calling a given model — not any single upstream's list price.

Code Router is a brand-new demo. The numbers, currencies, and payment options below are not finalized — every concrete amount on this page is a placeholder marked (TBD / 待补). Do not treat anything here as a committed price or available payment method.

Where to find it

Open the Pricing entry in the console navigation. You do not need to be signed in to view it; the table is public so callers can estimate cost before creating an API key.

The page renders one row per available model. Typical columns:

Model — the model name you pass in the model field of a request.
Type — how the model is billed (for example, per-token versus per-call).
Rate — the billing multiplier or unit price. (rate values TBD / 费率待补)
Group — which token/model group the rate belongs to (see below).

A screenshot of the live pricing table will be added once the demo UI is finalized. (screenshot placeholder)

How rates are expressed

Code Router prices most models per token, separating input (prompt) tokens from output (completion) tokens. Conceptually:

cost = input_tokens  × input_rate
     + output_tokens × output_rate

Rates are configured by the operator of the deployment, so the actual unit (per-token, per-1K, or per-1M tokens) and currency depend on how this instance is set up. (unit, currency, and all "per N tokens" numbers TBD / 计价单位与币种待补)

Some models may instead be billed per call (a flat charge per request, independent of token count). The Pricing table's Type column tells you which scheme applies to each model.

Model groups and token groups

Code Router uses groups to let one deployment offer different rates or model sets to different users:

Model groups

Bundle related models so a single rate or routing policy applies across them. Useful when several upstream channels back the "same" logical model.

Token groups

Each API key (token) can be assigned to a group. The group determines which rate table and which models that key can reach.

When you create an API key, the group it belongs to decides the prices you see applied in your usage records. If a model is not available to your group, calls to it are rejected before any charge is incurred.

Quotas

Independently of price, a deployment can cap usage with quotas:

Balance / remaining quota — how much spend is left on your account or on a specific key.
Daily limits — an optional ceiling on usage per day.
Per-key limits — a remaining-quota value attached to an individual API key, so you can hand out a key with a fixed budget.

A request is allowed only if it passes all applicable checks (group access, remaining balance, per-key remaining, and any daily limit). When any check fails, the gateway returns an error instead of forwarding the call upstream.

Topping up

Open your account / wallet area in the console.

Choose a top-up amount. (available amounts TBD / 充值面额待补)

Complete payment. (payment methods are not finalized for this demo — 支付方式待补)

Your balance updates and the added quota becomes spendable across your keys.

No specific payment method, currency, or amount is available or guaranteed in this demo. Treat the top-up flow above as a UI walkthrough only.

Reading your usage and cost

After you make calls, the console's usage/log views show per-request token counts and the charge applied, computed from the rate for that model and your group. Comparing those records against this Pricing page is the reliable way to confirm how a model is billed for your account.

Feedback

Pricing structure and figures for this demo are still being defined. If something looks wrong or you have a question, open an issue on the source repo: github.com/ssgzy/Router-new-api.